RCS Summer School 2024: Difference between revisions

From RCSWiki
Jump to navigation Jump to search
No edit summary
Line 122: Line 122:


== Sessions ==
== Sessions ==
{| class="wikitable"
{| class="wikitable table-left-aligned"
!Session
! width="20%" | Session
!Time and Location
! width="20%" | Time and Location
!Synopsis
! width="60%" | Synopsis
|-
|-
!
!
Line 281: Line 281:
* Understand how to automatically build, train, and tune the best machine learning models based on your data, while allowing you to maintain full control and visibility.
* Understand how to automatically build, train, and tune the best machine learning models based on your data, while allowing you to maintain full control and visibility.
* Get started with ML easily and quickly using pre-built solutions for common financial use cases and open source models from popular model zoos.
* Get started with ML easily and quickly using pre-built solutions for common financial use cases and open source models from popular model zoos.


'''Speaker:''' AWS
'''Speaker:''' AWS
'''Level:''' Introductory
'''Level:''' Introductory
'''Prerequisites:''' None
'''Prerequisites:''' None
|-
|-

Revision as of 22:22, 16 May 2024

Research Computing Services' 3rd annual summer school will run from Monday, June 10 through to Wednesday, June 12, 2024 from 9AM to 5PM. This summer school consists of various sessions and workshops throughout these 3 days and is completely free to all University of Calgary members.

Our goal for this year's summer school is to Empower our researchers: Inspiring what is possible on HPC infrastructure.

RCS Summer School 2024 Poster

Registration

Registration is required to attend the RCS Summer School sessions. Registration is free to all members of the University of Calgary.

Register now

There will be a limit of approximately 100 seats. If you are unable to attend after registering, please cancel/modify your registration or notify us via email.

Topics

  • Introduction to RCS services and HPC resources
  • Introduction to Linux & Bash command line
  • Using Linux utilities for large datasets
  • Hands on with Linux & Slurm: Workshop
  • Using Open OnDemand on ARC
  • Develop a research data management plan
  • Reproducible data management with Datalad
  • Digital File Management
  • Using containers in HPC with Apptainer
  • Managing scientific software with Conda
  • Research workflow development with Prefect
  • AWS: ML in the Cloud, a walkthrough followed by a workshop
  • NVIDIA: Workflow optimization using NVIDIA GPUs
  • Dell & AMD: Machine learning with Dell and AMD

Schedule

The summer school sessions will be held in ICT 102 and ICT 114. Refreshments will be available in ICT 114 on all 3 days.

Time June 10 June 11 June 12
Track 1 Track 2 Track 1 Track 2 Track 1 Track 2
8:30 AM Registration & check-in
ICT 102
Registration & check-in
ICT 102
Registration & check-in
ICT 102
9:00 AM Introduction to RCS
ICT 102, 9:00 AM - 9:20 AM
Jill Kowalchuk
Refreshments
ICT 114
The Alliance: Introduction
ICT 102
Brock Kahanyshyn
Refreshments
ICT 114
TBD

ICT 102

Refreshments
ICT 114
9:30 AM Introduction to Linux, Bash,
and the command line

ICT 102, 9:30 AM - 10:30 AM
Robert Fridman
Data in Motion: Navigating Storage Solutions for Active Research Data
ICT 114, 9:30 AM - 11:20 AM
Ian Percel
Introduction to HPC resources
ICT 102, 9:30 AM - 10:20 AM
Robert Fridman, Dave Schulz
Reproducible Data Management with Datalad: Part II
ICT 114, 9:30 AM - 10:20 AM
David Deepwell, Pedro Martinez
NVIDIA: Workflow Optimization with NVIDIA GPUs
ICT 102, 9:30 AM - 12:00 PM
Jonathan Dursi
10:00 AM Refreshments
ICT 114
10:30 AM Workshop: Hands on with Linux & Slurm
ICT 102, 10:30 AM - 11:50 AM
Robert Fridman
Linux tools & utilities for working with large data sets
ICT 102, 10:30 AM - 11:20 AM
Leo Leung, Dave Schulz
11:00 AM
11:30 AM Reproducible Data Management with Datalad: Part I
ICT 114, 11:30 AM - 12:20 AM
David Deepwell, Pedro Martinez
RCS Q&A period: Ask RCS anything
ICT 102, 11:30 AM - 12:00 PM
RCS Team
12:00 PM Open OnDemand on ARC
ICT 102, 12:00 AM - 12:20 AM
Leo Leung
Lunch break
12:00 PM - 1:00 PM
Lunch break
12:00 PM - 1:00 PM
12:30 PM Lunch break
12:30 PM - 1:30 PM
1:00 PM Research Data Management and Data File Management
ICT 102, 1:00 PM - 2:20 PM
Jennifer Abel, Alex Thistlewood, Ingrid Reiche
Refreshments
ICT 114
Dell & AMD: Machine learning with Dell & AMD
ICT 102, 1:00 PM - 1:50 PM

Rob Lucas

1:30 PM AWS: Inspiring the art of the possible
ICT 102, 1:30 PM - 1:50 PM

AWS

Refreshments
ICT 114
2:00 PM AWS: How AWS works with Researchers
ICT 102, 2:00 PM - 2:20 PM

AWS

2:30 PM AWS: Machine Learning with low-code workshop
ICT 102, 2:30 PM - 4:50 PM
AWS
Introduction to containers with Apptainer
ICT 102, 2:30 PM - 3:20 PM
Tannistha Nandi
Prefect for Research Workflow Development
ICT 102, 2:30 PM - 3:50 PM
David Deepwell, Pedro Martinez
3:00 PM
3:30 PM Managing scientific software with Conda
ICT 102, 3:30 PM - 4:20 PM
Dmitri Rozmanov
4:00 PM End of day: 4:00 PM
4:30 PM End of day: 4:30 PM
5:00 PM End of day: 5:00 PM

Sessions

Session Time and Location Synopsis

Introduction to RCS

9:00AM - 9:20AM

ICT 102

We will begin the summer school with a quick introduction by Jill Kowalchuk, the Interim director of Research Computing Services. We'll go through who RCS is and the services that we offer.

Speaker: Jill Kowalchuk Level: Introductory Prerequisites: None

Introduction to Linux, Bash, and the command line

9:30AM - 10:30AM

ICT 102

A quick crash course on how to use Linux, bash shell, and the command line in general. This beginner friendly session requires no prior experience to Linux. We recommend bringing your own device to follow along.

Speaker: Robert Fridman Level: Introductory Prerequisites: None

Workshop: Hands on with Linux & Slurm

10:30AM - 11:50 AM

ICT 102

A follow-up workshop that builds on the basics covered in the Linux introduction session and goes into depth on how to use Slurm, the scheduler that RCS uses in their high performance computing clusters. We recommend bringing your own device to follow along.

Speaker: Robert Fridman Level: Introductory Prerequisites: None

Open OnDemand on ARC

12:00 AM - 12:20 AM

ICT 102

Did you know you can run a Linux desktop on ARC? In this session, we will do a quick demo of ARC Open OnDemand, a web interface that allows users to submit jobs that need graphical user interfaces. We will also cover how to monitor your jobs through Open OnDemand.

Speaker: Leo Leung Level: Introductory Prerequisites: None

Data in Motion: Navigating Storage Solutions for Active Research Data

9:30AM - 11:20AM

ICT 114 Track 2

Planning for and requesting specialized storage for large research projects can be a daunting proposition. The variety of storage options and the expected justifications for allocations locally to UCalgary, at national supercomputing sites, and in the public cloud can quickly become overwhelming. This talk aims to provide an introduction to the cost/benefit tradeoff in using different storage systems, when to reach out to different support services around the university for help in making critical decisions, and basic techniques for providing a quantitative justification for a storage request.

Speaker: Ian Percel Level: Introductory Prerequisites: None

Reproducible Data Management with Datalad

June 10 10:30AM - 11:20AM

June 11 9:30AM - 10:20AM ICT 114

This workshop provides an introduction to digital data management with DataLad. Background content will be covered before conducting the primary hands-on training where attendees will create a small demonstrative research project containing data provenance.

Content to be covered includes: dataset basics, capturing data-provenance, and collaborative data analysis.

DataLad is a git-based version control system. Although no git knowledge is required, familiarity with git is strongly advised. Command line experience is required.

Speaker: David Deepwell and Pedro Martinez

Level: Introductory

Prerequisites: Command line experience

Introduction to HPC resources

9:30AM - 10:20AM

ICT 102

An introduction to high performance computing resources offered by RCS. We will go over how our infrastructure ties in to your research and how to make the most out of Slurm. How to download and transfer data with other institutions.

Speaker: Robert Fridman, Dave Schulz Level: Introductory Prerequisites: None

Linux tools & utilities for working with large data sets

10:30AM - 11:20AM

ICT 102

As researchers use larger and larger datasets, it is imperative to effectively handle and manage these datasets. In this session, we will go through some common methods to work with datasets using standard Linux tools and utilities. We will cover common use cases on how to download large datasets from the Internet, parsing text-based data using tools such as sed, awk, grep, and will then tie everything together with pipes.

Speaker: Robert Fridman, Dave Schulz Level: Introductory Prerequisites: Command line experience

RCS Q&A period: Ask RCS anything

11:30AM - 12:00PM

ICT 102

A general question and answers period where you can ask us anything related to RCS and HPC.

Speaker: The RCS team Level: Introductory Prerequisites: None

Research Data Management and Data File Management

1:00PM - 2:20PM

ICT 102

Managing your digital files and research materials is critical for keeping yourself organized, collaborating, and communicating with colleagues. In this session, we will cover Research Data Management (RDM) and Data Management Plan (DMP). We will also go over best practices in digital file management depending on your individual and organizational needs. This presentation will also discuss best practices, versioning, and how to document and share your file and folder convention using a README file.

Speaker: Jennifer Abel, Alex Thistlewood, and Ingrid Reiche (from The University of Calgary Libraries and Cultural Resources) Level: Introductory Prerequisites: None

Introduction to containers with Apptainer

2:30PM - 3:20PM

ICT 102

Make your research workflows reproducible through the power of containers. We will go through in detail how to run containers on ARC using Apptainer.


Speaker: Tannistha Nandi Level: Introductory Prerequisites: None

Managing scientific software with Conda

3:30PM - 4:20PM

ICT 102

Running customized scientific software on a shared HPC environment may be challenging. This session, we will go over how to set up customized software environments using Conda.

Speaker: Tannistha Nandi Level: Introductory Prerequisites: None

Prefect for Research Workflow Development

2:30PM - 3:50PM

ICT 102

Modernize your research workflows using Prefect, an open source workflow orchestration tool. In this session we will cover some of the fundamentals of building workflows with Prefect, with examples on how to deploy Prefect on local and distributed computing infrastructure.

Speaker: David Deepwell and Pedro Martinez Level: Introductory Prerequisites: None

AWS: Inspiring the art of the possible

1:30PM - 1:50PM

ICT 102

Learn what is possible on AWS Cloud for research.

Speaker: AWS Level: Introductory Prerequisites: None

AWS: How AWS works with Researchers

1:30PM - 1:50PM

ICT 102

AWS has many programs to support researchers such as credits, letter of supports, immersion days, working on proof of concepts. In this session, we will cover how we engage with researchers and what programs are out there to help accelerate your research with the AWS Cloud.

Speaker: AWS Level: Introductory Prerequisites: None

AWS: Machine learning with low-code workshop

1:30 PM - 4:45 PM

ICT 102

The Machine Learning (ML) journey requires continuous experimentation and rapid prototyping to be successful. In order to create highly accurate and performant models, data scientists have to first experiment with feature engineering, model selection and  optimization techniques. These processes are traditionally time consuming and expensive. In this workshop attendees will learn the following:
  • How the Low-Code ML capabilities found in Amazon SageMaker Data Wrangler, Autopilot and Jumpstart, make it easier to experiment faster and bring highly accurate models to production more quickly and efficiently
  • How to simplify the process of data preparation and feature engineering, and complete each step of the data preparation workflow
  • Understand how to automatically build, train, and tune the best machine learning models based on your data, while allowing you to maintain full control and visibility.
  • Get started with ML easily and quickly using pre-built solutions for common financial use cases and open source models from popular model zoos.

Speaker: AWS Level: Introductory Prerequisites: None

Workflow Optimization with NVIDIA GPUs

9:30AM - 12:20AM

ICT 102

We will discuss how to optimizing workflows with NVIDIA powered GPUs to help accelerate your research.

Speaker: Jonathan Dursi from NVIDIA Level: Introductory Prerequisites: None

Dell Presentation: TBD

1:00 PM - 1:50 PM

ICT 102

TBD

Speaker: Rob Lucas from Dell Level: Introductory Prerequisites: None